Back to skill

Security audit

Url Safety Triage

Security checks for vulnerabilities and agentic risk

Overview

The scanners report a URL safety triage skill that sends full URLs to a third-party service and uses a mutable CLI package, so ClawScan marks it suspicious due to privacy and supply-chain concerns.

Before installing, consider whether you are comfortable sending full URLs to the external provider, especially URLs containing private or sensitive information. The scanner also flags the unpinned `agentcash@latest` package; the verdict is suspicious, not malicious.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (7)

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 27)May include surrounding context.

md
| Task | Endpoint | Price |
|------|----------|-------|
| Full link verdict | `https://api.glyphcall.com/pack/url-safety?url=<full-url>` | $0.02 |
| Blocklist only | `https://api.glyphcall.com/phishing/check?url=<full-url>` | $0.01 |

Use the agentcash client so payment and retries are handled:

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 28)May include surrounding context.

md
| Task | Endpoint | Price |
|------|----------|-------|
| Full link verdict | `https://api.glyphcall.com/pack/url-safety?url=<full-url>` | $0.02 |
| Blocklist only | `https://api.glyphcall.com/phishing/check?url=<full-url>` | $0.01 |

Use the agentcash client so payment and retries are handled:

The workflow sends the full URL to a third-party service without a privacy warning.

Medium
Category
Output Handling
Confidence
98% confidence
Finding

The workflow instructs agents to submit complete URLs from emails, chats, and payment contexts to an external provider without warning about sensitive URL components. Such URLs can contain password-reset tokens, signed access credentials, private document identifiers, or personal data, and embedding them in a GET query can also expose them through request logging.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 33)May include surrounding context.

Use the agentcash client so payment and retries are handled:

sh
npx agentcash@latest check "https://api.glyphcall.com/pack/url-safety?url=https%3A%2F%2Fexample.com"
npx agentcash@latest fetch "https://api.glyphcall.com/pack/url-safety?url=https%3A%2F%2Fexample.com"

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 34)May include surrounding context.

sh
npx agentcash@latest check "https://api.glyphcall.com/pack/url-safety?url=https%3A%2F%2Fexample.com"
npx agentcash@latest fetch "https://api.glyphcall.com/pack/url-safety?url=https%3A%2F%2Fexample.com"

The result contains risk (low, medium, high), the blocklist matches per host, the redirect

MCP server referenced without pinned version: 'npx agentcash'.

Medium
Category
MCP Rug Pull
Confidence
98% confidence
Finding

Although this is a CLI invocation rather than an MCP server, agentcash@latest is a mutable dependency, not a pinned version. Running it through npx can download and execute a future compromised release with the agent's permissions, potentially exposing accessible credentials or payment-related resources.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx agentcash'.

Medium
Category
MCP Rug Pull
Confidence
98% confidence
Finding

The fetch command also executes the mutable agentcash@latest package, making the workflow dependent on whichever release is current at execution time. A compromised release could execute code with the agent's privileges; this is the same supply-chain weakness as the preceding command, not evidence of malicious intent.

Content

No source excerpt is available for this finding.